The Nuclear Decarbonization Option:
Profiles of Selected Advanced
Reactor Technologies
MARCH, 2012
A Clean Air Task Force Report
This document was prepared by the Clean Air Task Force for the Bipartisan Policy Center’s Energy Project. These
papers are meant to inform the project’s work on nuclear energy policy. The findings and recommendations offered in
these papers are those of the authors alone and not of the Clean Air Task Force, the Bipartisan Policy Center’s Energy
Project, the Bipartisan Policy Center, its founders, or its board of directors.
COVER IMAGE:
(c) philipus / www.fotosearch.com.
Clean Air Task Force is a nonprofit organization dedicated to reducing atmospheric
pollution through research, advocacy, and private sector collaboration.
MAIN OFFICE
OTHER LOCATIONS
18 Tremont Street
Beijing, China
Suite 530
Brunswick, ME
Boston, MA 02108
Carbondale, IL
617.624.0234
Columbus, OH
[email protected]
Washington, DC
www.catf.us
Houston, TX
This page intentionally left blank.
Table of Contents
Introduction 1
Mike Fowler
Director, Advanced Technology, Clean Air Task Force
Status of Small Modular Light Water Reactors in the US
3
Dr. Theodore U. Marston
Principal, Marston Consulting
High Temperature Gas Reactors
13
Andrew C. Kadak, PhD
Former Professor of the Practice, Massachusetts Institute of Technology
Breakthrough Nuclear Economics: One Possible Path –
the Floride Salt Cooled High Temperature Reactor (FHR)
21
Per F. Peterson, Tommy Cisneros, Mike Laufer, Raluca Scarlat, Nicholas Zweibaum
Department of Nuclear Engineering, University of California, Berkeley
Authors’ Biographies
31
The Nuclear Decarbonization Option:
Profiles of Selected Advanced Reactor Technologies
Introduction
WE LIVE IN A WORLD divided by many issues, but most
policy-makers accept the basic premise that increasing
the availability of affordable low-carbon energy would
make the world healthier, wealthier, and safer. Conventional fuel delivery systems are strained in many regions,
the global geopolitics of energy supply are fraught, and
carbon dioxide emissions, despite decades of debate since
Rio and Kyoto, are rising faster now than at any point in
history. And still, billions remain without regular access
to electricity and mobility.
Nuclear energy provides more than 40 percent of all
low-carbon electricity generated in the world today. That
contribution could grow, but public perceptions of safety
remain a key challenge—particularly post-Fukushima—
and competitive costs, as always, will be paramount. In
order to assess the impact that advanced technologies
could play in the development and deployment of new
nuclear reactor designs, the Clean Air Task Force asked
several national leaders in nuclear technology to give us
their perspectives on key policy-relevant issues.
We asked Dr. Ted Marston, former Chief Technology
Officer of the Electric Power Research Institute, to write
for us on small, modular light water reactors (smLWRs).
Dr. Andrew Kadak, former Professor of the Practice in
Nuclear Engineering at the Massachusetts Institute of
Technology, examines the prospects for high-temperature
gas-cooled reactors (HTGRs). And Dr. Per Peterson, Chair
of the Nuclear Engineering Department at University of
California, Berkeley, explores the future of some fluoride
molten salt reactors (called FHRs).
Their conclusions are important and offer reasons for
optimism:
l Small, modular light water reactors (smLWRs):
With modest development efforts, smLWRs, using fuel
and systems quite similar to modern LWRs, could
offer significantly enhanced safety over the existing
nuclear fleet, deployment flexibility (e.g., staged investment and repurposing of some existing infrastructure),
and potential cost-reductions through efficiencies of
factory manufacturing.
l High-temperature gas-cooled reactors (HTGRs): HTGRs, using extremely heat-resistant, encapsulated fuel (already demonstrated in the United States
and elsewhere) offer the possibility of nearly meltdownproof reactors, higher thermal efficiencies, and expanded uses for nuclear energy (e.g., manufacturing of
zero-carbon liquid transportation fuels), as well as many
of the potential deployment and manufacturing advantages of smLWRs.
l Fluoride salt-cooled High temperature Reactors (FHRs): And FHRs, using the same heat-resistant, encapsulated fuel as HTGRs, but with coolants of
dense molten salt compounds, could retain many of
the advantages of HTGRs at a greatly reduced size,
offering the potential for breakthrough economics if
designs prove out.
For a world struggling to reduce carbon emissions while
sustaining and increasing economic growth, and understandably concerned about the potential risks of nuclear
energy, the advantages these advanced reactor designs
offer could be profound. But bringing these concepts to
commercial reality will require sustained development,
especially for the more advanced concepts. Our hope is
that these papers will help to inform the debate about how
governments and the private sector should support that
development.
This report does not aspire to cover the full scope of
potentially important nuclear power technologies. Korean
and Russian firms are developing smLWRs that could be
important in some markets, and technologies that address
the life-cycle of nuclear fuel and waste—including fast
neutron reactors and thorium-based reactors—also could
be important. More radical designs, such as the liquid
fluoride thorium reactors (developed by Oak Ridge National Laboratory to use liquid fuels, rather than solid),
may be able to provide even more dramatic advantages
on safety, cost, and fuel cycle issues. Sub-critical reactors
driven by particle accelerators may one day be able to
convert low-value nuclear materials directly into energy.
We will explore the potential of these technologies in
future reports.
Mike Fowler
Director, Advanced Technology
Clean Air Task Force
IN T RODU C T I ON 1
This page intentionally left blank.
THEODORE U. MARSTON, PHD.
Principal, Marston Consulting
Status of Small Modular Light Water Reactors in the US
Introduction to Small Modular Reactors
Small modular reactor (SMR) is a relatively new term for
the old concept of small (less than 350 MWe output) reactors that are principally manufactured in a shop environment and shipped to the plant site for assembly. At last
count, the International Atomic Energy Agency lists over
65 SMRs under development. Most of these are in early
stage design, i.e. conceptual design, often called ‘paper
reactors’. This white paper focuses on the type of SMR
that has the most complete designs and which appears
closest to deployment, at least in the US. These SMRs are
cooled and moderated with light (that is, ordinary) water,
similar to the operating nuclear power plant fleet in the
US. In this white paper we call this type of SMR the small,
modular, light water reactor (smLWR).
l Smaller, older coal-fired power plants in US face in-
creasing pressure from environmental requirements,
including potential carbon constraints. Some estimate
as much as 45GWe of old coal plants will be closed in
the next 5 to 10 years.
As a result of the utility interest in smLWRs, there is an
increasing interest from the traditional and non-traditional US nuclear plant suppliers. NuScale Power, a start-up
company in Oregon, was the first firm to seriously pursue
smLWR business and associated licensing issues. Babcock
and Wilcox, a 150 year old boiler manufacturer, soon followed. More recently, Westinghouse has entered the
competition. The primary reason for their interest is the
potential nuclear plant sales, but there are a number of
ancillary factors. From a public policy perspective these
include:
Potential Benefits of smLWRs
l The smLWR could help restore the US leadership
Energy demand and economic benefits
of smLWRs
l Provides for US manufacturing jobs
Since 2008, there has been a growing interest in the US
in the development and deployment of small modular
light water reactors (smLWRs). There are a number of
reasons for this interest. For the US utilities, these include:
l US utilities need new baseload generation capacity due
to insufficient reserve margins in many of the North
America Electricity Reliability Corporation (NERC)
regions.
l Most US electric utilities cannot afford to take the financial risk of building a large advanced light water
reactor (ALWR) because the project represents a major portion of the entire capitalization of the average
nuclear utility.
lMost US electric utilities cannot accommodate a
1300MWe or greater block of new generation in their
service territories. Additionally, they wish not to put
that much generation on ‘one shaft’, i.e. too great a generation risk if the plant is unavailable.
in nuclear energy
l Helps restore the US nuclear supply chain
infrastructure
l Opportunity to sell US shop-assembled modules
overseas
l Revise (improve) the timeliness of the US regulatory
process for new technology
l Increased profitability from exploiting the ‘learning
curve’ for smLWRs
Both the House of Representatives and the Senate have
prepared bills to support the development and deployment
of small modular reactors. The motivation behind the
Congressional bills is to stimulate job growth and the
economy. Estimates of the economic benefit1 of one small
Note: Prior to the Fukushima accident, it was thought that the
smLWRs could replace retired coal-fired generation in key locations.
The impact of the accident on the acceptability of nuclear power to
the public makes this option more uncertain.
M ARST ON / smLWR s 3
smLWR plant include:
l 7000 jobs
event. The smLWRs have independence of the modules in a multi-module plant. The containments are
located below grade and are designed to sustain high
internal pressures and protect equipment from external threats.
l $1.3 B in sales
l $0.63B in value-add
l $0.4B in earnings
l $35M in business taxes
President Obama’s administration has issued broad national emission reduction goals and specific guidance for
federal installations; these are:
1)National greenhouse gas (CO2) emissions must be
reduced by 83% below 2005 by 2050, 42% by 2035
and 30% by 2025.2
2)All federal installations must reduce their greenhouse
3
gas (CO2) emissions by 28% by 2020.
Deployment of smLWRs at DOE and DoD facilities,
particularly national laboratories would support these
emission reduction goals for federal facilities.
Safety Benefits for smLWRs
The smLWR concept has a number of attractive features
from the perspective of power plant safety and protection
of the public health and safety from radioactive releases.
This section presents some generic considerations that
set the smLWR apart from traditional LWR designs. This
is not to say that the current designs are not safe. Industry efforts at continuous improvement, compliance with
NRC regulations, and the passing of countless inspections
by the NRC assure the safety of the US LWR fleet. These
features include:
l Greater chance for ‘inherent safety’ – the general de-
signs incorporate concepts of natural circulation of the
primary system following reactor scram,4 cooling water provided from large tanks or reservoirs using gravity, decay heat rejection systems that are mostly inside
the reactor containment structure, no need for gridsupplied or on-site emergency power for extended
periods of time, large battery capacities to run instrumentation and control (I&C) systems and limited valve
operations, digital I&C systems (which provide more
resilient reactor control than analog), and extended
time for operators to develop robust recovery strategies. These attributes are similar to those of the larger, passive advanced light water reactors (ALWRs),
such as the Westinghouse AP1000.
5
l Smaller source term for severe accidents (e.g., those
that can lead to core damage) and more effective decay heat removal – these reduce the likelihood of a core
damage event and reduce the amount of radioactive
release in the low probability event of a core damage
4
M A R S T O N / smLWR s
l Integral designs (described in more detail below) sig-
nificantly reduce risks of large break loss-of-coolant
accidents (LOCAs) – there are no large main coolant
pipes to break. The largest credible break size is a few
square inches. Reactor pressure vessel (RPV) vessel
penetrations are limited by design. All reactor coolant
pumps (RCPs) use canned rotor pumps (which have
no seals that could result in leaks).
l Larger primary inventory basis and pressurizer – The
primary system cooling water inventory is roughly three
times that of conventional LWRs on a MWth basis. The
relative pressurizer volume of smLWRs is also much
larger, which improves ability to maneuver the reactor
through some transients.
l Vessel layout facilitates natural convection – the smL-
WR vessels are very tall and small diameter in comparison with conventional LWRs. The entire secondary
system is well above the core in elevation. One of the
smLWR designs run normally on natural circulation.
Those with RCPs are designed to transition smoothly
to natural circulation, if the RCPs fail to operate.
l Below grade containment has safety and security ad-
vantages – All of the US design smLWRs have containments located below grade, which make aircraft impact
an incredible event. The used fuel pools are similarly
located below grade and some are inside of containment. Flooding in below grade spaces must be considered carefully in the design of the smLWRs to assure
the protection of vital equipment if the plant is subjected to flooding conditions.
l All of the smLWRs utilize a probabilistic safety analy-
sis in the design process to assure the likelihood of core
damage frequency (CDF) is as low as practical. The estimated CDFs and radioactive release frequencies for
the smLWRs are projected to be equal to or lower than
the ALWRs, which are substantially below those of the
operating light water reactors.
Status of smLWR Development
Three smLWR designs appear to have the greatest potential for commercial success in the 2020 timeframe. The
designs are integral, pressurized water reactors (PWRs),
i.e. designs in which the major nuclear steam supply system (NSSS) components, including reactor and core, steam
FIGURE 1
Shows the reference 540 MWe,
12 Module NuScale Plant with the Details of the
Containment and Pressure Vessel Structure
generators, pressurizer and pumps (if part of the design),
are housed in a single pressure vessel. The original integral
PWR was designed to power the German commercial ship,
N.S. Otto Hahn, commissioned in 1968, which sailed
without incident for over one million kilometers. The ship
was converted to conventional power in 1979 for economic reasons.
The three designs selected are:
l NuScale reactor (45 MWe, natural circulation) under
development by NuScale Power, Inc.,
l mPower reactor (180 MWe, forced circulation) under
development by the team led by the Babcock & Wilcox
Company,
l SMR (>225 MWe, forced circulation) under develop-
ment by Westinghouse Electric Company LLC.
We briefly summarize each design, provide a figure
showing the overall plant, the containment structure and
the integral reactor pressure vessel, and identify the commercialization strategy. Following the individual smLWR
descriptions, a table compares and contrasts the three
designs. Each of the designs uses a shortened length variant of standard commercial 17X17 PWR fuel. All of the
three smLWRs are designed for 60 years of operation.
NuScale
6
NuScale Power, Inc., originally a privately held company
in Corvallis, Oregon USA, is designing and commercializing, a modular, 45 MWe Pressurized Water Reactor.
Each NuScale module has its own combined containment
vessel and reactor system, and its own designated turbinegenerator set. NuScale power plants are scalable, allowing
for a single facility to have just one or up to 12 units. In a
multi-module plant, one unit can be taken out of service
without affecting the operation of the others. The reference
NuScale plant has 12 modules and generates 540 MWe.
Each NuScale plant component is modular and is designed for fabrication in a number of existing facilities in
the USA and around the world. In theory, the construction
should be less complex, lead times shorter, and costs more
predictable and controllable. The NuScale containment
containing the reactor pressure vessel measures approximately 60 feet in length and 14 feet in diameter. All components are transportable by barge, truck or rail.
The NuScale design requires no operator actions in the
first 72 hours after plant shutdown. The ultimate heat sink
is a pool of water internal to the reactor building which
sits below grade. This pool is sized to accept decay heat
from all modules without boiling for the first 72 hours. If
pool or reactor cooling is not restored or other sources of
make-up to the pool are not provided the reactor pool will
slowly boil off over time allowing the decay heat load to
subside. By the time the pool inventory is exhausted,
decay heat will have dropped to a point that air cooling
on the shell of the containment is sufficient to maintain
the reactor cores cooled and covered.
Each module has its own independent turbine generator set that is also small enough to be delivered as a
complete, modular package. In addition to the modularity of the design, the plant will be constructed using
modular construction techniques.
Favorable characteristics of the NuScale design include:
l Small size of the Reactor Building
l Significant portion of the building underground
l Limited access into reactor building
l Passive safety systems
l Limited vital equipment
7
l Ultimate heat sink internal to the reactor building
l Smaller fission product inventory per reactor
l Use of remote and automated technologies
l Risk informed approach to design
M ARST ON / smLWR s 5
mPower
The Babcock & Wilcox (B&W) Company builds small light
water reactors for the US Navy. Originally, their commercial smLWR was a modular, 50 MWe, natural circulation, pressurized water reactor planned for emergency
power use by the US Air Force as part of their Small
Military Power Plant (SMPP) program. This reactor was
called the B&W Gem50. In 2009, B&W uprated the Gen50
to 125 MWe using forced circulation with internal pumps
and in 2011 the power increased to 180 MWe.
The new reactor concept is called the B&W mPower
reactor. Some of the key mPower features include all forgings capable of being fabricated in the US; transportable
by rail; use of “proven” technology, wherever possible;
and licensing features that the NRC would find “comfortable”. The technology which is unproven will be tested
extensively. The internal control rod drives , required by
the design, are the novel principal technology in the
mPower design. B&W formed a joint venture LLC with
Bechtel Power in 2010
The mPower team began pre-application interactions
with the US Nuclear Regulatory Commission (USNRC) in
July 2009. Design certification application for the mPower design is expected for either the 4th quarter of CY 2012
or 1st quarter of CY 2013.
The lead plant project for the mPower design is a joint
mPower – Tennessee Valley Authority (TVA) proposed 2
unit power plant on the TVA Clinch River Site. This proposed station will provide emission-free electricity to the
FIGURE 2
Illustrates the mPower 360
MWe Reference Plant with 2 Modules,
Below Grade Containment Details
and Integral Pressure Vessel
6
M A R S T O N / smLWR s
Oak Ridge National Laboratory (ORNL) to help the laboratory meet the previously discussed CO2 reduction targets
mandated by the Obama Administration.
The mPower safety approach incorporates:
l Passive safety
– No safety related emergency diesel generators
– No core uncovering during design basis accident
conditions
l 2 trains of emergency core cooling systems (ECCS)
l Natural circulation assured when the RCPs are
shutdown
l Multi-level Digital I&C systems with diverse analog
shutdown system
l In containment reactor water storage tank
l Full pressure ECCS heat exchanger inside of
containment
B&W is:
l One of the last major nuclear component manufacturers left in the US. In the 1970s, there were five major
component manufacturers
l The major nuclear heavy component supplier to the
US Navy
l Capable to make the entire integral mPower vessel in
house.
l The mPower vessel is designed to be rail-transportable
throughout the US.
The outstanding testing to be completed includes:
lComponents
– Canned rotor (sealess) reactor coolant pumps
– Control rod drive mechanisms (CRDMs)
FIGURE 3
Illustrates the Reference Westinghouse 225 MWe SMR Plant with one module, Details of the Below Grade
Containment and Integral Pressure Vessel
– Fuel mechanical testing
8
– CRDM/fuel integrated test
– Fuel critical heat flux
– Emergency high pressure condenser
l B&W has constructed an integrated systems testing
facility in Bedford, VA. Among other things, the
facility will investigate:
– Heat transfer phenomena
– Steam generators performance
– Loss of coolant accident response
– Pressurizer performance
– Reactor control
Westinghouse Small Modular Reactor
Westinghouse, a wholly owned subsidiary of Toshiba,
announced in February of 2011, they would be developing
a 225MWe small modular reactor. All primary components
of the Westinghouse SMR, a 250 MWe integral pressurized
water reactor, are located in the reactor vessel. It features
passive safety systems and a reduced number of components compared to other reactor designs. Many of the
reactor’s parts can be manufactured in factories off-site
and delivered by truck or rail to the plant site.
The Small Modular Reactor enables Westinghouse to
integrate all that they know about operating nuclear plants,
designing and licensing plants, and passive safety into a
small design that will provide additional nuclear power
options for their owner/operator customers. Much of the
Westinghouse SMR’s design draws from the company’s
AP600 and AP1000 designs.
Some of the safety and security features of the Westinghouse SMR include:
l In the event of a station blackout, the Westinghouse
SMR is designed to shut down automatically, requiring no intervention for seven days, using gravity and
natural cooling.
l Spent fuel pool is located below grade and contains
heavily reinforced concrete structures lined with steel.
l Spent fuel will remain covered with gravity flow of water from safety related Decay Heat Removal Tanks for
a period of seven days
l Containment designed for protection against seismic
events, natural disasters, and aircraft impact.
Westinghouse SMR design certification document (DCD)
package is on-schedule for NRC submittal in 2012. Component and system tests and planning are on schedule.
The testing program underway for the CRDMs should
confirm the applicability of the proven AP600 and AP1000
CRDMs in the Westinghouse SMR environment.
M ARST ON / smLWR s 7
Summary Table of smLWRs
The following summary table compares and contrasts the three smLWR designs. The information in the table is provided by the NSSS designers. The decay heat removal time listed is the minimum time without any intervention. One
design can theoretically go indefinitely with no intervention. The others require minimal intervention, such as filling
of unpressurized tanks and pools, to maintain adequate decay heat removal for an indefinite period of time.
9
smLWR Summary Table
Challenges to Commercialization
The smLWRs have a number of challenges to successful
commercialization. These include regulatory and economic challenges and overcoming the negative public
reaction to the Fukushima accident. There are a number
of regulatory issues related to the licensing and deployment of smLWRs in spite of the fact that the current licensing base is built around light water reactors. The NRC
regulations are designed for large plants. All of the PWRs
in the US are non-integral, i.e. separate reactor pressure
vessel, steam generators, reactor coolant pumps and pressurizers. Some of the major issues under consideration
by the US NRC and the principal contentions are presented below:
l Nuclear insurance – Price Anderson - The premium
payments paid annually by the owner/operators for
Price Anderson liability insurance are based on the
number of reactors owned. The premium is based on
large LWRs. It is believed that the second party liability incurred by smLWRs are significantly less than for
plants with core thermal ratings as 30 times greater.
Proposals to amend the premium structure to account
for core power output are under development.
8
M A R S T O N / smLWR s
l Annual fees – The NRC assesses their annual fees to
the licensees on the basis of the number of plants. The
NRC has to collect 90% of their annual operating budget from direct fees charged for review, licensing and
inspection or indirect fees charged to each nuclear power plant. In 2010, each nuclear plant was charged
$4,719,000 regardless of the power output. It is felt
that each smLWR should not be charged the same
amount as the larger plants. A sliding scale is proposed
to deal with the size discrepancies.
l Staffing – Current control room staffing requirements
are based on large reactors with fully analog control
room technology. The control rooms and I&C systems
for the smLWRs should be fully digital, possibly with
a separate analog system to provide redundancy and
diversity in the shutdown of the smLWRs. The inherent safety of the new smLWR designs in conjunction
with the fully digital control systems with a high degree of automation should permit the safe operation
of the smLWRs without the tradition one control team
for each reactor, used in the existing plants. Alternative staffing requirements are under discussion.
l Security – Security requirements for US LWRs have
increased substantially since the terrorist events of 11
Sept 2001. The requirements are based on new threats
and the ability for existing reactors to respond to those
threats. The smLWR designs include security in the
design and have taken major steps to reduce the security needs. For example, the entire nuclear steam supply system (NSSS), spent fuel pool and containment
for all designs are located below grade. The access to
control and radioactive material areas is significantly
reduced over the existing plants. State of the art security and intrusion detection systems are part of the design. Therefore, it is believed that adequate security of
a smLWR can be maintained with simplified security
requirements. Proposed simplifications are under development for smLWRs.
l Emergency planning – size of emergency planning
zones – The emergency planning and the zone of evacuation for US plants is based on the existing fleet. The
smLWRs are significantly different in terms of source
term in the case of a core melt event. The smLWR core
damage frequencies are orders of magnitude lower than
10
what is required in the NRC regulations. The containments are located below grade and the long term cooling needs of a beyond design basis core damage event
are much less. For these reasons, the industry believes
the current emergency planning zones and notification
requirements can be greatly simplified and still protect
the health and safety of the public. Proposed simplifications of emergency planning for the smLWRs are
currently under development. Such simplification is
required to locate a smLWR near regions of high populations, such as those surrounding the existing coal
plants that will likely be shut down. This simplification
will be a major challenge in light of the 2011 Fukushima accident in Japan.
Regulatory challenges could make smLWRs noncompetitive. If the licensing of smLWRs become protracted affairs, the attractiveness of such small plants will
vanish. The best hope for smLWRs to be competitive lies
in the assumption that they can be licensed, built and
commissioned quickly.
The primary economic challenge to the commercialization of smLWRs is whether the electricity production costs
are (1) affordable and (2) competitive with other forms of
generation. With regard to affordability, smLWRs offer
potential optionality to the US electric utilities, when the
only real options for large generation additions are gas
fired, coal fired or large nuclear plants. SmLWRs, being
smaller and modular, potentially offer a more manageable
nuclear option. SmLWRs are more ‘affordable’, i.e. less of
a fiscal risk. They can be deployed in much smaller increments, matching the utilities’ load growths better and
reduce the ‘single shaft’ generation risk to an acceptable
level.
Competing with other forms of electricity generation
is a much greater challenge today. Vast amounts of natural gas are being discovered across the US in so-called
tight gas (shale) deposits, resulting in cheap and abundant
natural gas. The current spot market price of natural
gas is less than $3.00/MMBTU. Carbon restraints (taxes
or credits), which would improve the competitiveness of
smLWRs, appear unlikely to arise in the near future.
However it is expected that carbon emissions from large
stationary sources will be reduced systematically over time
one way or another, and US utilities are very interested
in reducing their ‘carbon footprints’. If the economics of
the smLWRs are what some of the designs claim, there is
a real chance to compete with natural gas fired plants,
particularly when carbon constraints are in place. The cost
competitiveness of smLWR depend heavily on achieving
the following opportunities:
l Streamline design and manufacturing are necessary to
offset the economies of scale of other generation options, particularly nuclear plants. ALWRs are becoming larger and larger due to the economies of scale. The
only prospect to reverse this effect for the smaller smLWRs is to streamline the shop fabrication of the NSSS
and other modules, ship them to the site and install
them rapidly. The requisite quality standards must be
maintained throughout the entire process.
l Modularity of the smLWRs provides the opportunity
to transform how we design, build, operate and decommission nuclear power plants.
l Reduce construction time by modularization and construction efficiencies
l SMRs do not require loan guarantees. This sets the
smLWR apart from the larger ALWR, which currently
benefit from federal loan guarantees, especially for regulated utilities. Experience shows the loan guarantee
process to be a protracted and expensive affair, requiring the expenditure of significant political and fiscal
capital.
How the impacts of the Fukushima accident affect
smLWR development and deployment is unclear. The
passive nature of the safety systems and the reduced need
for AC power following shutdown should be positive attributes. Likewise, the depth of the containment should
mitigate certain security concerns, but may raise flooding
concerns. However, the idea of locating a number, up to
M ARST ON / smLWR s 9
twelve, of smLWRs at a single plant site may become a
liability in the eyes of the public. The sequential failure of
the Fukushima reactors followed by the hydrogen explosions will be long lasting memories for the public. It may
be difficult to convince the public that more reactors at a
site is safe, in spite of the fact that the single reactor failure source term is much smaller than current reactors and
that there is little chance for system interaction in the new
designs.
process initiated with a funding opportunity announcement (FOA). A draft of the FOA is out for comment and
the final FOA is expected to issue in 2Q 2012. In addition
to the FOA on smLWR design and certification, the DOE
expects to fund R&D related to smLWRs on the following
subjects:
l Regulatory issues
l Safety analysis, particularly from a probabilistic safe-
ty analysis point of view
l Fabrication efficiency improvements
US Market Potential for smLWRs
l Digital instrumentation and control.
The potential smLWR market in the US is quite large,
with three primary opportunities for deployment. The
nearest term opportunity is to build the initial plants to
provide low emission electricity to DOE and DoD facilities
that are subject to the Executive Order. This opportunity
will be part of the current DOE SMR Program. The second
opportunity for smLWR deployment is to provide baseload
(or near baseload) generation capacity resulting from
general load growth. Baseload generation has not been
installed to any degree since the 1970s and 1980s in the
US. Most of the recent generation additions have been
combined cycle turbines fueled by natural gas. The net
demand for the US will grow by 30% in the next 20 years,
however. There are renewable portfolio standards in 30
states in the US, so some of this growth will be met with
renewable generation, but others, like Ohio, have clean
energy standards which include the traditional renewables
and other forms of non-emitting generation, such as
nuclear plants and large dams. Even with renewables and
some gas, a role for nuclear could emerge due to general
demand growth.
Each of the smLWR designers has an extensive test
program to meet the regulatory and reliability needs of
the new design.
The third opportunity is replacement of existing fossil
baseload generation. The DOE estimates that 75 GWe of
old coal (operating life > 35 years) could be retired by
2025. In addition to the old coal plants, there is almost
20GWe of old, inefficient natural gas and oil-fired boilers
in the US that are over 50 years old. The potential smLWR
replacement percentage of this retiring generation may
be as high as 10% or about 10 GWe of deployment.
Research, Development and
Demonstration Needs
The US Department of Energy has adopted the small
modular reactor as one of their keystone programs. One
of the major elements of the DOE SMR Program is support
of the design and licensing of two smLWRs with up to
$425M. The designs will be selected through a competitive
10
M A R S T O N / smLWR s
Conclusions
SmLWRs offer a potentially attractive nuclear option to
the US electricity new build landscape that will most
likely be dominated by natural gas. However, the actual
deployment of smLWRs will depend on two factors. The
first factor is the simplification of emergency planning
zones (EPZs) for the smLWRs by the NRC and the need
for public acceptance of small, very safe nuclear plants
near to population centers. This is a necessary, but not
sufficient condition for success deployment. The smLWRs
must make economic sense to the owner/operators. They
must be affordable, financeable and the levelized cost of
electricity must be competitive with other low or no emission generation technologies.
Footnotes
1 Solan, David, Economic Impacts of Small Modular Reactors: Considerations and Recent Events, Center for Advanced Energy Studies, The
Idaho National Laboratory, Platts 2nd Annual Small Modular Reactors
Conference, May 2011, Washington, DC.
2 White House November 2009.
3 White House October 5, 2009. Executive Order 13514.
4 A reactor SCRAM is a sudden, complete, shutdown of the reactor fission power production by the insertion of control rods or other mechanisms.
5 The ‘source term’ is the amount of radioactivity that potentially could be
released to the environment following an accident.
6 The Fluor Corporation obtained a majority position in NuScale Power,
LLC in October 2011.
7 The ultimate heat sink is the final cooling reservoir or medium for decay
heat from a shutdown reactor.
8 Although similar to fuel for conventional LWRs in many ways, some
structural and mechanical differences necessitated by the geometry of
this reactor core necessitate new testing and validation.
9
Table Notes: NA – not available, FOAK – first of a kind, NOAK – nth of a
kind, Passive safety – no active components required for emergency
core cooling, No&L – number and active length of fuel assemblies,
OTSG – once through steam generator, Part 52 – 10CFRPart52 (onestep licensing process), Part 50 – 10CFRPart50 (conventional licensing
process), Hybrid 50/52 – First plant licensed under Part50, followed by
certification of design under Part52, vacuum – containment is operated
at sub-atmospheric pressures to improve insulation and response to
pressurization, rail ship – major components shippable by rail.
10 For an example of a preliminary analysis of these issues, see, Probabilistic Risk Assessment of NuScale Reactor During the Design Phase,
Welter et al, 2009.
M ARST ON / smLWR s 11
This page intentionally left blank.
ANDREW C. KADAK, PH.D.
Former Professor of the Practice
Massachusetts Institute of Technology
High Temperature Gas Reactors
Introduction
High temperature gas reactors (HTGR) are gaining increasing interest globally due to their high thermal efficiencies for electricity production, high temperatures for
process heat applications, and inherent safety. The Fukushima nuclear accident in Japan has also renewed interest
1
in more accident tolerant reactors. Germany was the
historic leader in the development and application of
HTGRs, centered around their Jeulich Research Institute,
2
which used pebble bed reactor technology. In the United
States, General Atomics was the leader in HTGR technology using helium cooled prismatic-block reactors, several of which were built in the 1960s and 1970s. Japan
has an operating prismatic-block helium cooled-research
reactor in operation since 1999, with a goal of demonstrating hydrogen production, and South Africa was about to
build a 165 Mwe direct cycle pebble bed reactor (PBMR)
until the economy of the country and rising costs forced
its cancelation. More recently, China has taken on significant leadership in some aspects of HTGR technology.
In the United States, Exelon, the largest nuclear generating company, was a 10% owner of the South African
pebble bed project and was planning to build the US version of the PBMR until low natural gas prices and management changes within the company resulted in Exelon
pulling out of the project. Exelon went as far as submitting
pre-licensing white papers to the Nuclear Regulatory
3
Commission before cancelling the project. The 2005
Energy Policy Act mandated the construction of the Next
Generation Nuclear Plant (NGNP) by 2021 which was to
be a high temperature gas reactor to provide heat for
hydrogen production at the Idaho National Laboratory.
Three competing design teams made up of such companies
as Westinghouse Electric Company, General Atomics,
AREVA and Shaw Engineering representing both pebble
and prismatic reactors were preparing conceptual designs
4
for the NGNP. The NGNP industry alliance recently an-
nounced that they have decided to support the construction of the AREVA prismatic version of the NGNP for
process heat production since Westinghouse (PBMR) and
General Atomics withdrew from the competition. Unfortunately, this project was recently relegated to a research
and development project instead of the construction of a
demonstration plant by the Department of Energy.
More positively, China has taken the lead using pebble
bed reactor technology having an operational 10 Mwth
research reactor in operation since 2000 at Tsinghua
University’s Institute of New and Nuclear Energy Technology (INET) near Beijing. Construction of a full scale 200
Mwe electric generating station using this technology
(High Temperature gas cooled Reactor – Pebble-bed
Module – HTR-PM) for Huaneng Group is underway near
Shidao in Shandong Province. The Chinese have been
developing the high temperature pebble bed technology
since the early 1990s. The HTR-10 reactor has been used
to test fundamental helium technology systems including
valves, online refueling systems, fuel manufacture and
qualification in Russian test reactors, steam generator
design and overall operations. This reactor ran several
significant tests including loss of helium flow, turbine
trips, loss of offsite power supply, and helium blower trips
without insertion of shutdown rods. The Chinese have
used this experience to design their full scale prototype
power reactor. In addition they have built impressive test
facilities at their research facility outside of Beijing to
perform testing of key components of the HTR- PM demonstration reactor. They have built a 10 Mw helium test
loop, a scaled version of their helical once through steam
generator to verify performance of the design; a full scope
test facility of the fuel handling system; and test facilities
to confirm the design of their control rod drive mechanism
and shutdown system. As these facilities demonstrate,
Tsinghua University’s Institute of Nuclear Energy Technology has taken the lead in high temperature reactor
development.
KADAK / H T G R 13
FIGURE 1
Sites of HTR-10 on the left, and New HTR-PM Construction in China on the right
In terms of licensing, the HTR-PM has passed the licensing review with over 2000 questions that were raised by
the Chinese regulatory authorities prior to the issuance
of a construction permit. The Chinese National Nuclear
Safety Administration has licensed the French 1000 Mwe
PWR, the Russian VVER, Sodium Cooled Fast Reactor,
the Westinghouse AP-1000 and the French EPR in the
past and is viewed as a qualified licensing authority supported by universities and science and engineering institutes in China. China’s regulations are patterned after US
high level regulations but are less prescriptive.
The promise of high temperature reactors has always
been safety and the ability of producing heat at between
750 C to 950 C which are temperatures impossible to attain for light water reactors. This high temperature is
possible due to the use of silicon carbide coated fuel
particles containing minute amounts of uranium, graphite and using the inert gas, helium, as a coolant. This high
temperature offers opportunities not only for 50% more
efficient electric production but also for replacing CO2emitting fossil fuels such as coal, oil and natural gas in
the production of high quality steam. The temperatures
produced are high enough for chemical production of
5
hydrogen , and for such activities as oil sands production
(avoiding the use of coal or natural gas to extract the oil),
oil refining, conversion of coal to synthetic fuels, and
extraction of oil from shale. While these latter missions
may seem counter to reducing CO2 emissions, the need
for fossil fuels for transportation remain and the use of
high temperature gas reactors to avoid fossil emissions
for their extraction could be an important avoidance
strategy.
14
K A D A K / H TG R
Technology Basics
Gas cooled reactors are not new. The earliest demonstration of high-temperature, helium-cooled reactors was the
Dragon plant in the UK which started operation in 1964.
The first gas cooled reactors were used to produce steam
in a Rankine cycle, as a conventional light water reactor
might produce, but at temperatures in the 700 C range
(e.g., at the Peach Bottom and Fort St. Vrain reactors in
the US). More recent designs which go to higher temperatures use a Brayton cycle with gas turbines to produce
electricity. The gas turbine cycle allows for even higher
thermal efficiencies of up to 50%, reducing cooling needs
by 50%.
For electric power production numerous options exist:
a direct cycle takes the hot helium directly into a gas
turbine; an indirect cycle which takes the heat from the
reactor to an intermediate heat exchanger to make steam
or another gas system to power gas turbines. This intermediate heat exchanger allows for a multiplicity of end
uses for the energy of the reactor produces making the
reactor essentially a universal heat engine.
There are two major types of high temperature gas reactors: Prismatic-block, and pebble bed. Both use the same
type of fuel form, however – poppy seed-sized kernels of
uranium coated with several thin layers of protective
material including silicon carbide and graphite, called
6
“TRISO.” These tiny particles are shaped into a pebble
with additional graphite or, alternatively, a chalk-sized
compact which is inserted into a graphite block as shown
on Figure 2. In the pebble form each pebble, which is approximately the size of a tennis ball, contains 10,000 of
the TRISO particles.
FIGURE 2
Prismatic Fuel Blocks and Pebble Fuel Using TRISO Fuel Particles
Prismatic HTGRs
7
The prismatic reactor developed by General Atomics is
based on the 40 Mwe Peach Bottom gas reactor which
operated from 1967 to 1974 in Pennsylvania and the 330
Mwe Fort St. Vrain gas reactor, which operated intermittently from 1979 to 1989 in Colorado.8 General Atomics
developed the Gas Turbine Modular High Temperature
Gas Reactor (GT-MHTGR) as a candidate for the NGNP
effort as a direct-cycle helium-cooled reactor, shown in
Figure 3 below. This plant is capable of producing approximately 300 Mwe. The reactor shown on the right
consists of stacking the fuel blocks above 10 high making
up the core which needs to be refueled every 18 months
similar to light water reactors.
Pebble Bed HTGRs
Pebble bed reactors were developed in Germany over 30
9
years ago. At the Juelich Research Center, the AVR
pebble bed research reactor rated at 40 Mwth and 15 Mwe
operated for 22 years demonstrating that this technology
works. A steam generator was used to generate electricity
through a conventional steam electric plant. Germany also
built a 300 Mwe version of the pebble bed reactor – the
Thorium High Temperature Reactor (THTR) - but it suffered some early mechanical and political problems that
10
eventually lead to its shutdown. In December of 2000,
the Institute of Nuclear Energy Technology of Tsinghua
University in Beijing China, achieved first criticality of
11
their 10 Mwth- 4Mwe pebble bed research reactor. In
12
the Netherlands, the Petten Research Institute has developed pebble bed reactors for industrial applications in
the range of 30 Mwth.
concluded to be about 250 Mw thermal or about 110 Mwe
to allow for rapid and modular construction as well as
maintaining its inherent safety features. The key inherent
safety feature is the ability to remove heat from the vessel
without need for emergency core cooling systems as are
found in light water reactors. Thus to maintain this feature
the cores diameter must be small enough to allow for the
heat transfer to occur without overheating the fuel. Additionally, the control rods are located outside of the core
which necessitates a smaller diameter core to be sure that
the nuclear reaction can be controlled. A pebble bed reactor is graphically illustrated in Figure 4. The reactor core
contains approximately 360,000 uranium fueled pebbles
about the size of tennis balls. Each pebble contains 9 grams
of low enriched uranium in 10,000 tiny grains of sand-like
microsphere coated particles each with its own a hard
silicon carbide shell. The unique feature of pebble bed
FIGURE 3
General Atomics Prismatic GTHMR – 300 Mwe
Advances in basic reactor and helium gas turbine technology have produced a new version of the pebble bed
reactor concepts. The optimum size for a pebble bed was
KADAK / H T G R 15
FIGURE 4
Graphic of a Pebble Bed Reactor (left) and Actual Pebbles in HTR-10 Reactor
Showing Graphite Block Reflector into which Pebbles are Loaded
reactors is the online refueling capability in which the
pebbles are recirculated with checks on integrity and
consumption of uranium. This system allows new fuel to
be inserted during operation and used fuel to be discharged
and stored on site for the life of the plant. This online
refueling capability allows for more efficient use of the
fuel since 24 months of fuel does not have to be loaded
all at once and minimizes the excess reactivity (potential
for fission) that needs to be managed to assure shutdown.
Additionally, before the pebbles are discharged, they are
measured for the degree of utilization (burnup) and can
be reinserted if there is still sufficient fuel in for another
pass through the core. This capability not only maximizes the fuel utilization, it also does not require shutdown
for refueling which takes away from power generation
improving the economics of the plant.
It is projected that each pebble will pass through the
reactor 6-10 times before discharge in a three year period
on average. Due to the on-line refueling capability, plant
maintenance outages are only required every 6 years. The
internals are made of carbon blocks which act as a neutron
reflector and structural support for the pebble bed.
“Space” Frame HTGRs
A more modern high temperature gas reactor conceptual
design configuration is the Galvin Energy Module 100
Mwe plant shown on Figure 5 below. The unique feature
of this design is that it is an indirect Brayton cycle using
helium gas on both sides of the heat exchanger allowing
the nuclear plant to be a true heat engine.
16
K A D A K / H TG R
Figure 5 also shows another unique feature of the design
on the power conversion system. Shown are “space” frames
which are “lego” like structures which are factory manufactured and shipped to the site by truck or train allowing
for rapid assembly to improve it competitive advantage
allowing for completion of construction in less than three
years.13,14
The HTGR Safety Case
The basis for the safety of high temperature reactors is
founded on three principles. The first is the very low
power density of the reactor which means that the amount
of energy and heat produced is volumetrically low and
that there are natural mechanisms such as conductive and
radiative heat transfer that will remove the decay heat
from the reactor even if no active core cooling is provided.
This is fundamentally different than light water reactors
which still require an active coolant flow (even if passively generated) to remove decay heat. By limiting the
size of both the prismatic and pebble bed cores, all of the
decay heat can be passively removed from the reactor
vessel even upon total loss of all cooling. This is significant
since the peak temperature that is reached is roughly 1600
C which is far below the 2000 C fuel damage temperature
of the silicon carbide particles and the 2865 C melting
temperature of uranium dioxide fuel. Additionally, it takes
about 70 to 80 hours to reach the peak temperature after
which the temperature begins to drop due to the reduction
in decay heat generated in the core.
The long time it takes to heat up the core is due to the
large heat capacity of the graphite blocks that may up the
reflector. The conclusion that decay heat cannot melt an
HTGR core is supported by tests and analysis performed
in Germany, Japan, China, South Africa and the US. Both
the low power density and the relatively small size of the
core allow for the unique safety of this reactor technology.
The second key principle is that the reactor naturally
shuts itself down as the temperature goes up (due to impact of the temperature on core reactivity). This feature
(shutting down upon increase is temperature) is a feature
which light water reactors do not possess and prevents
the type of accident that occurred at the Chernobyl nuclear plant. A demonstration of an intrinsic shutdown of an
HTGR reactor took place at the HTR-10 reactor near
Beijing before an expert panel from the International
15
Atomic Energy Agency in 2004.
The third principle is that the silicon carbide - which
forms the tiny containment for each of the 10,000 “TRISO”
fuel particles in a pebble – can retain its fission products
under all anticipated reactor conditions, provided adequate
standards have been met in its manufacture. In tests
performed to date on fuel reliability, it has been shown
that micospheres can be routinely manufactured with
initial defects of less than 1 in 10,000. In safety analyses,
it is assumed therefore that 1 in 10,000 of these microspheres has a defect that would release the fission products
into the coolant. Since the amount of fuel in each particle
is very small, only 0.0007 grams, then even with this assumption and under accident conditions, the release from
the core would be so low that no offsite emergency plan
would be required. In essence, it is recognized that the
FIGURE 5
fuel cannot be made perfectly, but the plant is still safe
because is has natural safety features that prevent meltdowns. Manufacturing high-quality fuel is a key factor in
the safety of high temperature gas reactors, however.
A safety issue that needs to be addressed with all graphite reactors is that of water and air ingress. Water ingress
adds positive reactivity to the core, but this is offset by
the negative temperature coefficients. Each have been
analyzed and can be effectively addressed in the design.
For air ingress accidents, at high temperatures, oxygen
reacts with carbon to form CO and CO2. This oxidation
and corrosion of the graphite is a both an exothermic and
endothermic reaction depending upon the conditions.
Analyses and tests in Germany have shown that it is very
difficult to “burn” the graphite in the traditional sense,
but it can be corroded and consumed. The key issue for
the pebble bed reactor is the amount of air available in
the core from the reactor cavity and whether a chimney
can form allowing for a flow of air to the graphite internal
structure and fuel balls. Tests and analyses have shown
that at these temperatures the natural circulation required
for “burning” is not likely due to resistance to natural
circulation flow. The high-temperature corrosion process
in an HTGR core suffering from air ingress would be
similar to a slow diffusion process which allows for mitigating actions should an air ingress accident occur.
HTGR Economics
No matter what the environmental, public health, safety
and energy security advantages nuclear energy may
offer, if the product is not competitive, it will not be used.
GEM 100 Pebble Bed Reactor Showing Modularity Features
Helium to Helium Heat Exchangers (2)
Space Frame Power Conversion System
– gas turbine
Pebble Bed Reactor
KADAK / H T G R 17
High temperature gas reactors, in order to maintain their
inherent safety characteristics as outlined in the safety
case, are smaller in power output (100 to 300 Mwe) than
light water reactors with outputs of 1,500 Mwe or larger.
Thus, in order to compete against the economies of scale,
they must compete on economies of mass production.
Due to the smaller size, modularity is possible reducing
construction time and cost based on mass production of
components. A typical 1000 Mwe or larger will cost upwards of $ 8 Billion each and take approximately 5 years
to build compared with around $500 Million for about a
16
100 Mwe modular plant with a 3 year construction time.
There is rarely a need for 1000 Mwe of new power all at
once in the US. Smaller modular reactors allow for a util17
ity to plan its new generation based on needs. Should
there be a need for a 1,000 Mwe plant, 10 modules could
be built at the same site. The concept calls for a single
control operating all 10 units through an advanced control
system employing many of the multi-plant lessons of
modern gas fired power plants in terms on modularity
and automatic operation. A unique feature of this modularity approach is that it allows one to generate income
during construction as opposed to paying interest during
construction
The appropriate metric for competition is the total cost
of power and not just the overnight capital cost to build
the plant which the numbers above represent. Additional
costs that must be factored include interest on the cost of
capital, operating and maintenance costs and fuel costs.
The total cost as represented by a cost per kilowatt-hour
is how electricity is sold to consumers. When compared
to larger plants or other forms of generation, high temperature gas reactors need to be within 10% or so of the
competition to be considered by generating companies.
Nuclear energy struggles due to high capital costs. However, over the long run with volatile fuel prices that are
the major cost element of fossil fueled power stations,
nuclear plants have been shown to be the least cost long
term solution on a production cost basis. Clearly, if the
cost of carbon is included in energy prices, nuclear’s competitive advantage will be strengthened even further.
The only active project to price the cost of high temperature gas reactors is the Chinese HTR-PM. The HTRPM consists of twin pebble bed reactors that provide heat
to steam generators that drive a single 200 Mwe steam
turbine. Recent estimates place the cost of this plant at
about $ 2,000/kwe excluding R&D and infrastructure cost
to transmit the power. The Chinese have procured substantial portions of the plant in preparation for construction such that this estimate is based on actual numbers.
18
K A D A K / H TG R
While the economics of Chinese production and US
production are quite different, even if that number was
doubled, it would still be in the competitive range of current LWRs being proposed in the US.
Conclusions
High temperature gas reactors offer the potential not only
of a more accident tolerant reactor system but a nuclear
heat engine that can provide electricity and process heat
applications in a more efficient and CO2 free manner.
While the Chinese are leading the way with their pebble
bed power module presently being built, the future direction in high temperature gas technology is with more efficient helium gas turbines. As the US struggles with its
own energy policy direction, researchers at the Idaho
National Laboratory, Oak Ridge National Laboratory and
the nations of the European Union are continuing to
develop the technology for the future.
Footnotes
1http://www.nytimes.com/2011/03/25/business/energyenvironment/25chinanuke.html?_r=1.
2 In a pebble bed reactor, fuel is contained in spherical elements that
move slowly through the core over time. In a prismatic block reactor,
the fuel is held in immovable structural elements, more like a conventional reactor core.
3http://pbadupws.nrc.gov/docs/ML0624/ML062400070.pdf.
4 http://www.ngnpalliance.org/ February 15, 2012 Announcement.
5 For example, using chemical processes such as sulfur-iodine, hybrid
sulfur and high-temperature electrolysis processes.
6 “TRISO” means ‘tristructural-isotropic.’
7http://gt-mhr.ga.com/description.php.
8 While the basic high temperature gas reactor technology was shown
to be satisfactory, there were problems with the graphite reflector
cracking and mechanical problems with the seals of the helium circulators. Due to these problems, Ft. Saint Vrain achieved very low utilization during this period.
9 AVR – Experimental High Temperature Reactor: 21 Years of Successful Operation for a Future Energy Technology, Association of German
Engineers (VDI) – the Society for Energy Technologies, Dusseldorf,
1990.
10 The THTR was a very large pebble bed reactor that was not properly
designed in that the control rods, instead of being inserted in the outer
reflector region where inserted directly into the pebble bed causing
the pebbles to crack. Additionally there were problems with the insulation in the piping systems which came loose during operation which
required repair. The final blow to THTR was the Chernobyl accident
which turned public sentiment against nuclear power and the utility
chose to shut it down after a brief period of operation.
11 Z.Zhang, Y. Sun, “Economic Potential of modular Reactor Nuclear
Power Plants based on the Chinese HTR-PM Project”, Nuclear
Engineering and Design, 2007.
12 Petten ACACIA reference. http://www.iaea.org/inisnkm/nkm/aws/
htgr/fulltext/htr2004_d15.pdf.
13 Berte, M.V., “Modularity in Design of the MIT Pebble Bed Reactor,”
Massachusetts Institute of Technology S.M. Thesis, Department of
Nuclear Engineering, 2004.
14 Hanlon-Hyssong, Jamie, “Modularity of the MIT Pebble Bed Reactor
for use by the commercial power industry”, Massachusetts Institute
of Technology S.M. Thesis, Department of Nuclear Engineering, 2008.
15 Dong, Yujie, “Status of Development and Deployment Scheme of
HTR-PM in the People’s Republic of China”, Institute of Nuclear and
New Energy Technology, Tsinghua University, Beijing, China, Interregional Workshop on Advanced Nuclear Reactor Technology for Near
Term Deployment, Vienna, Austria, July 4-8, 2011.
16 “High Temperature Gas-Cooled Reactor Projected Markets and
Preliminary Economics”, INL/EXT-10-19037.
17 Benouaich, Michael, “Option approach to investment in modular
nuclear electricity-generating capacity”, Massachusetts Institute of
Technology S.M. Thesis, Department of Civil and Environmental
Engineering, 2002.0-19037, August 2010.
KADAK / H T G R 19
This page intentionally left blank.
PER F. PETERSON, TOMMY CISNEROS, MIKE LAUFER,
RALUCA SCARLAT, NICHOLAS ZWEIBAUM
Department of Nuclear Engineering
University of California, Berkeley
Breakthrough Nuclear Economics: One Possible Path –
the Floride Salt Cooled High Temperature Reactor (FHR)
Introduction
Today nuclear power has the lowest production costs of
any major energy source in the United States. This cost,
which includes operations and maintenance (O&M), fuel
and waste management, now averages 2.1 cents per
kilowatt-hour, compared to over 3 cents for coal [1]. As a
result the current economics of nuclear energy favor work
to extend the operating life of existing reactors. The construction of significant numbers of new reactors has
proven to be more problematic, however, due to their high
construction costs and the difficulty U.S. utilities face in
financing large construction projects that can approach
or exceed $10 billion. While the development in the near
term of small modular light water reactors (smLWRs) can
be expected to reduce the difficulty of financing new nuclear construction projects, construction costs exceeding
$4000 per kilowatt are likely to continue to suppress the
commercial demand for new nuclear construction as long
as carbon-adjusted costs for coal and gas remain low.
Over the past century the capital costs of fossil power
stations have been reduced by two major trends that have
not occurred for nuclear power. One has been a general
trend, enabled by continued advances in improved high
temperature materials, toward significantly higher operating temperatures that enable increased efficiency in the
production of electricity. These improvements in thermal
efficiency enable the design of more compact and less
expensive power plant systems. The second trend has been
to move away from pure steam power cycles to power
plant designs that use a combination of a gas turbine
producing around 60% of the total power, with a smaller
steam turbine “bottoming cycle.”
Gas turbines are radically different from these earlier
steam turbines, since they had originally been developed,
and successfully deployed, for the purpose of lifting air-
planes into the air (something difficult to image that a
steam locomotive or a cargo ship steam turbine could do).
The compact design of these early “aero-derived” gas
turbines and the high efficiency of the combined cycle
system (reaching as high as 60%) resulted in a significant
reduction in capital cost. Conversely, during this same
period increasingly stringent requirements for pollution
control have resulted in a competing trend of capital cost
escalation for fossil power stations.
For commercial nuclear power stations, no comparable
significant increases in efficiency have occurred in the last
half century. The modern light-water cooled reactors
(LWRs) being built today have power conversion efficiencies at most a few percentage points higher than the low
32% efficiency of LWRs that entered construction in the
1970’s. Moreover, in analogy to the increased fossil construction costs due to new pollution control requirements,
the evolutionary reactor designs that emerged in the 1980’s
and 1990’s after the 1979 Three Mile Island accident had
significant increases in systems, steel, and concrete compared to the earlier reactors, to provide additional redundancy and physical separation of active safety systems.
These additional safety systems in evolutionary designs
resulted in increased construction costs. This trend has
now reversed with the introduction of passive safety systems in U.S. LWR designs such as the Westinghouse AP1000 and General Electric ESBWR, as well as smLWRs,
1
but capital costs still remain high.
Because combined cycle fossil plants have much lower
capital costs than older, conventional steam plants, a
natural question for nuclear power is whether any similar
technical changes might have the potential to provide a
similar reduction in nuclear capital costs, while maintaining the low O&M, fuel, and waste management costs of
current LWRs.
PET ERSON / F HRs 21
FIGURE 1
Molten salt reactor technology
(top) was initially developed for use on
nuclear-powered bombers (lower left).
The Aircraft Reactor Experiment was
designed, built and in 1954 was tested
at ORNL (lower right) for 1000 hours,
delivering 2.5 MWth of power at a
temperature of 860°C (1500°F) [2, 3].
Further reductions in LWR capital costs are likely possible through improved designs and construction methods.
But fuel and coolant used in LWRs creates intrinsic limitations in power conversion efficiency. One approach to
select among alternative fuels and coolants is to seek
analogies to the technical characteristics that have reduced
construction costs in fossil technologies. The closest analogy to the aero-derived gas turbines that enabled the
development of combined cycle fossil plants would be a
class of reactors studied from the 1950’s through the 1970’s
for application to nuclear-powered aircraft, and then to
stationary power plants. The early Aircraft Nuclear Propulsion (ANP) and subsequent Molten Salt Breeder Reactor (MSBR) programs used fluoride salts as coolants and
as solvents for their fuels, resulting in reactor designs with
liquid fuels that operated at low pressure and high temperature.
22
P E T E R S O N / FHRs
During this period the U.S. built and successfully operated two molten salt reactors. The first, the Aircraft Reactor Experiment (ARE) shown in Fig. 1, achieved criticality in 1954 and provided a ground-based demonstration
of a sufficiently lightweight, high temperature reactor to
power the jet engines of strategic bombers. While the ARE
proved that reactors sufficiently lightweight to power
aircraft are technically viable, the launch of the Sputnik
satellite in 1957 shifted U.S. nuclear deterrence strategy
and development investment toward missiles and the ANP
program was cancelled in 1961. So the major emphasis
driving nuclear power development remained with the
Navy’s submarine propulsion program, where the strict
weight performance limits for aircraft did not exist, and
where with the time pressure to deploy naval reactors
rapidly light water emerged as the preferred reactor coolant, due to the ability to adapt conventional steam-cycle
equipment to nuclear service.
As described in the Appendix, after the cancellation of
the ANP program work to develop a Molten Salt Breeder
Reactor (MSBR) for commercial applications continued
at Oak Ridge National Laboratory (ORNL), in parallel
with a substantially larger program to develop uraniumfueled liquid metal fast breeder reactors (LMFBR) centered
at Argonne National Laboratory. Extensive research,
documented in many thousands of pages of technical
reports now available on-line [4], addressed problems in
neutronics, heat transfer, chemistry, materials, and component design, leading to the construction and operation
of the 8-MWth Molten Salt Reactor Experiment, which
ran for 26,000 hours between 1965 and 1969. As competition for resources increased, ultimately the molten salt
reactor was cancelled in 1972, so that resources could be
concentrated into the development of the LMFBR.
Starting from ideas proposed three decades later in
2002 [5], today a consortium of universities (UCB, MIT,
and UW Madison), partnered with ORNL, are leading a
new DOE-funded Integrated Research Project (IRP) to
study a new high temperature reactor technology. The
Advisory Panel that guides this IRP is chaired by a recently retired Chief Technology Officer of Westinghouse
and has membership including recently retired Chief
Nuclear Officer from Entergy, a recently retired Nuclear
Science and Technology Division Director from ORNL,
and representatives from two major nuclear technology
consulting firms.
These new reactors – which are currently only conceptual – would combine fuels derived from high temperature
gas cooled reactors (HTGRs) with fluoride salts studied
in the ANP and MSBR programs as their coolant. This
combination of fuel and coolant, shown in Fig. 1, provides
unique attributes. Called fluoride salt cooled high-temperature reactors (FHRs), the use of ceramic fuel and core
structures in these reactors gives them the capability to
deliver heat at temperatures above 600°C (1050°F), but
without the high pressures of LWRs and HTGRs. Thus
FHRs can achieve high power conversion efficiency while
having similar compact size and low mass as did their
original nuclear aircraft ancestor.
The entry barriers for any new commercial reactor
technology are large. The baseline FHR design (solid,
high-temperature ceramic fuel with a fluoride salt coolant,
rather than a liquid fuel) has been selected to enable
predictable and high reliability, as well as licenseability
within a commercially attractive time frame.
2.0 Key Economic Metrics for FHRs
Experts who estimate the capital cost of new engineered
facilities like nuclear reactors focus their efforts on quan-
tifying the volumes and masses of materials needed for
construction, counting numbers of components, estimating labor requirements, determining schedules, and
identifying sources of risk and uncertainty in these materials, labor, and schedule estimates. Using cost data from
earlier projects, and adjusting for interest during construction, cost estimators can then generate total top-down
cost estimates that can be used to assemble bids to perform
new projects. Large FOAK projects tend to attract a small
number of suppliers, who must build large contingencies
into their FOAK prices due to the large downside risk
of cost overruns. This is why nearly all modern Nuclear
Steam Supply System (NSSS) reactor vendors have
trended towards becoming as vertically-integrated as possible, oftentimes in a transnational sense.
The risk associated with FOAK construction of large
reactors reinforces the arguments for developing small
modular reactors (SMRs), where the financial risks associated with building a FOAK unit are reduced. With
SMRs one can expect larger numbers of companies to
enter into the competition to construct FOAK power plants
and accept risk in order to gain experience and the ability to bid accurately for subsequent projects. Because the
SMR market is likely to attract more suppliers, customers
can have higher confidence that the prices they pay are
close to the actual costs of construction. Moreover, customers can benefit from easier financing for SMR projects
and from the flexibility that SMRs provide in adjusting
total generation capacity to optimal levels.
2.1 Metrics for Comparing Capital Costs of
Different Reactors
Some key nuclear plant design metrics that affect capital
cost include:
l Reactor building height
l Reactor building volume per MWe
l Reactor building concrete volume per MWe
l Steel mass per MWe
l Core inlet/outlet temperature
l Primary coolant cost per MWe
l Core power density
l Primary system volume per MWe
l Primary system metal mass per MWe
Comparisons of values of these metrics can provide a
rough, preliminary means to assess the potential differences in construction costs of different types of reactors.
Likewise, these metrics, along with other metrics relating
to technical risk and financing, can be used to guide design
decisions during conceptual design.
PET ERSON / F HRs 23
The earliest FHR economics study, performed in 2004
[6], used the reactor primary system volume and building
volume as the primary construction cost metrics. In this
study an FHR was designed that used the same reactor
building and reactor vessel size as the 1000-MWth, 380MWe S-PRISM liquid metal reactor (LMR, also known as
the Integral Fast Reactor, or IFR, developed after the
cancellation of the U.S. LMFBR program). While the
primary system and reactor building volume of this early
FHR design was the same as the IFR, when designed with
a core power density of 10 MWth/m3 (approximately twice
the typical power density of a HTGR), could produce 2400
MWth (1145 MWe), 2.4 times more power than the IFR.
With further scaling for power conversion system costs,
this 2004 study concluded that the capital cost of the FHR
would be approximately 55% that of the IFR. This study
also developed a capital cost comparison with the 600MWth, 286-MWe GT-MHR, concluding that the capital
cost would be 61% of this HTGR’s cost. Subsequent research has suggested that the optimal FHR power density is likely to be yet higher, between 15 to 30 MWth/m3,
so an FHR of this physical size could have a power output
of 3600 to 7200 MWth (1700 to 3400 MWe), and thus
yet lower capital cost.
While very large FHR power levels are possible in principal, to control technical risk early commercial deployment would involve smaller units with lower power levels.
Thus the next major FHR design study completed in 2008
[7], shown in Fig. 2, was a 900-MWth, 410-MWe design
that operates with an approximate core power density of
16 MWth/m3. For this new FHR design the reactor vessel
volume, per MWe, is 21% of the IFR vessel volume, almost
a factor of 5 more compact.
While it is difficult to directly compare primary system
costs between FHRs and ALWRs, due to their very different construction methods, the building volumes for these
reactors can be estimated relatively easily and compared.
FIGURE 2
The 900-MWth, 410-MWe PB-AHTR
point design developed in 2008 by the
U.C. Berkeley Nuclear Engineering Department [7].
24
P E T E R S O N / FHRs
Table 1 compares building volumes for the 2008 FHR
design and various advanced LWRs (and HTGRs), showing that the FHR buildings tend to be significantly more
compact.
Building height is another key metric for capital cost.
The height of a reactor building results from the “stacking”
of key reactor components, along with the crane-bay height
needed to provide lift space above the reactor. FHRs use
natural circulation of their coolant to remove decay heat,
but fortunately because the fluoride salts have very effective natural circulation ability (due largely to their high
volumetric heat capacity) relatively short heights are
needed to drive heat removal with reasonably low temperature differences in the loops. Because construction is
a vertical process, building height affects construction
time. The 2008 FHR design shown in Fig. 2 has a total
height, from base-mat to roof, of 35 m (115 ft). This is
approximately half of the 75 m (240 ft) height of typical
LWR and HTGR buildings, and supports a conclusion that
FHR construction times can be shorter than for these
other reactor technologies.
2.2 FOAK Costs and Risk Management
While it is clearly possible to reduce the capital cost of
new reactor designs by introducing new materials and
technologies, the decisions for where to take on technical
risk must be made strategically, because they will contribute to the first of a kind (FOAK) costs and time needed to
develop a new reactor design. Fortunately, the methods
used to license passive safety systems for LWRs are generally applicable to FHRs as well, and during the last
decade the U.S. has made substantial investments in the
development of fuel and materials for HTGR and LMR
technologies that can also be applied to FHRs.
New fuels are commonly agreed to be one of the most
challenging and time-consuming areas for reactor technology development due to the cost and long time periods
TABLE 1
Comparison of building volumes for various ALWRs and HTGRs, with an FHR [7].
Reactor
Power
(MWe)
Reactor &
Auxiliaries
Volume
(m3/MWe)
Turbine Building
Volume
(m3/MWe)
Ancillary
Structures
Volume
(m3/MWe)
Total Building
Volume
(m3/MWe)
1970’s PWR
1000
129
161
46
336
ABWR
1380
211
252
23
486
ESBWR
1550†
132†
166
45
343
EPR
1600
228
107
87
422
GT-MHR
286
388
0
24
412
PBMR
170
1015
0
270
1285
Modular PB-AHTR
410
98
104
40
242
Reactor Type
†
The ESBWR power and reactor building volume are updated values based on the Design Certification application arrangement drawings.
needed to develop, irradiate and examine new fuels. Fortuitously for the FHR, the U.S. Next Generation Nuclear
Plant (NGNP) project has established the full set of capabilities to fabricate, irradiate, and examine the coated
particle fuels needed for FHRs [8]. This NGNP fuels
program has demonstrated excellent fuel performance,
with no fuel particle failures occurring. Because FHRs
operate at much higher power density than HTGRs, their
fuels reach full depletion more rapidly, allowing FHR fuels
to be developed and tested on an accelerated schedule.
High temperature materials are a second key area where
development and qualification is challenging. Here again
the U.S. has made significant investments in the last decade
to improve the understanding of materials performance
at high temperatures, both for applications to HTGRs as
well as LMRs [9]. A major issue for these high temperature
reactor designs, and for the FHR, is that they use metallic
structural materials at sufficiently high temperatures for
long-term creep deformation to be important. It is challenging to design vessels, pipes and heat exchangers to
operate under conditions where such time-dependent
behavior occurs. The DOE has sponsored significant work
to update the ASME Boiler and Pressure Vessel code to
integrate high temperature design rules into a new section
called Division 5 [10]. Large databases of materials properties are needed to predict time-dependent creep phenomena. For this reason, the FHR design has focused on
using code-qualified materials such as 316 stainless steel,
Alloy 800H, and Alloy N that already have extensive
property databases. Division 5 also includes design rules
for graphite that are used in FHRs for key core structures
including radial reflectors and upper core support structures. While there are multiple vendors today who have
the ability to design and construct components from these
materials for nuclear applications today, further work to
confirm the corrosion resistance of these materials under
specific FHR conditions remains to be done and is the
major work of UW Madison in the IRP effort.
3.0 FHR Safety, Design and Licensing
FHRs have negative fuel and coolant temperature reactivity feedback that causes the fission process to shut
down if the reactor overheats, as well as additional diverse
and redundant reactivity control and shutdown systems.
As with other reactors, a key safety requirement for FHRs
following shutdown is the removal of decay heat generated by fission projects that continue to undergo radioactive decay after the fission process is stopped.
In FHRs decay heat is removed by buoyancy-driven
natural circulation of the salt coolant. Under normal
shutdown conditions decay heat is removed by a reliable
normal shutdown cooling system shown in Fig. 3 that can
function in both active (with AC power) and passive (no
AC power) heat removal modes. Under emergency conditions where the normal shutdown cooling system is not
available, a fully passive, safety-related Direct Reactor
Auxiliary Cooling System (DRACS) is available that can
PET ERSON / F HRs 25
FIGURE 3
Flow diagram for a typical FHR primary system and intermediate loop.
reject decay heat to ambient air without any AC power
indefinitely. These systems are designed to maintain all
primary loop metallic structures, including the reactor
vessel and intermediate heat exchangers, below temperatures where accelerated creep or other structural damage
could occur. For beyond design basis accidents where both
the normal shutdown and DRACS decay heat removal
systems fail, the FHR design allows decay heat to be rejected to the reactor cavity at a rate sufficient to maintain
fuel temperatures below failure thresholds, even under
conditions where the metallic structures in the primary
system may have been damaged significantly.
With this diverse and redundant collection of decay heat
removal paths, the peak temperature of FHR fuel during
design basis transients and accidents remains more than
500°C (900°F) below the 1600°C threshold where fuel
damage and release of fission products can occur. This is
a major difference compared to LWRs, HTGRs, and LMRs,
where the reactor power limits are established by fuel
failure limits at the peak local temperatures reached in
2
the core during design basis accidents. Predicting the
peak local temperature in a reactor core is challenging,
and requires careful identification of phenomena that can
affect local temperatures, development and validation of
computer models for predicting peak temperature, and
26
P E T E R S O N / FHRs
assessment of uncertainty in the predictions. Conversely
for FHRs, the fuel safety margin is so large that localized
fuel failure is deterministically impossible during design
basis transients and accidents, and instead one is faced
with the simpler challenge of predicting the averaged core
behavior that determines the peak coolant outlet temperature and the temperatures reached by metallic structures.
The transient response of FHRs during design basis
events generally involves a subset of phenomena that occur in passive LWR designs, since with the very high
boiling temperature of the salts heat transport occurs only
with single-phase flow. Thus FHR transient phenomena
can be modeled using the same codes as used in the licensing of LWRs [11]. Fortuitously the convective heat transfer phenomena that occur in FHRs can be replicated in
inexpensive reduced-scale experiments using heat transfer oils [12]. This allows FHR safety codes to be validated
against a large base of thermal hydraulics test data at
reasonable cost, with experiments using actual salts providing confirmatory data. Additionally, simulant fluids
give substantial flexibility to the reactor designer in performing proof-of-principle experiments for key design
modifications or assessing the viability of novel components.
FHRs also have some attributes that could increase their
costs and thus must be addressed carefully. These include
the need to understand and control overcooling transients
to prevent and mitigate consequences from salt freezing,
the need to adapt the conventional beryllium industrial
safety programs used by other industries to be integrated
with the FHR’s radiation safety program, the need to
develop and demonstrate methods to enrich lithium using
3
non-mercury-based technologies, and the need to identify and select a preferred power conversion technology
from the options of open-air-Brayton cycles, supercritical
steam cycles, and closed gas Brayton cycles.
Because licensing by the U.S. Nuclear Regulatory Commission is a key gateway activity for the commercial deployment of any new reactor technology in the United
States, the MIT/UCB/UW IRP conducted a workshop with
20 licensing and reactor safety experts to review how licensing methods developed in the U.S. for LWR, HTGR
and LMR technologies can be adapted to use for FHRs
[13]. This workshop identified no major issues for adapting existing licensing methods to FHRs.
A second key gateway activity for commercial deployment of FHRs is the construction of a 10 to 20 MWth
Fluoride salt-cooled High-temperature Test Reactor
(FHTR). While a large number of test reactors have been
built and operated worldwide to provide the basis for
understanding LWR, HTGR, LMR and MSBR technologies, information from these earlier and current test reactors does not provide the information on the integrated
performance of FHR systems, particularly for start up,
power operation, and shutdown transients, needed to
assess performance and safety characteristics of FHRs
sufficiently to support licensing of a commercial prototype.
University-based materials corrosion testing, irradiation,
and thermal hydraulics experiments will be needed to
support the design and licensing of the FHTR. Additionally, FHTR fuel development and testing using existing
NGNP capabilities will be needed, as well as design, construction and operation of an FHR Component Test Facility for non-nuclear testing of FHTR components under
prototypical chemical and thermal conditions.
A final key gateway activity for commercial deployment
of FHRs is the demonstration of reliable operation and
the capability to achieve high levels of availability. Here
the major elements involve the systematic identification
of potential materials and component degradation mechanisms; research to understand these mechanisms; particularly corrosion of structural materials in the presence
of prototypical salt chemical conditions; the development
of reliable component designs and methods to perform
online monitoring, in-service inspection, maintenance
and replacement; the use of event trees and probabilistic
risk assessment to quantify overall system reliability; and
the validation of these models through operating experience in the FHR Component Test Facility and the FHTR.
4.0 Conclusions
The FHR concept derives from a combination of a coolant,
fuel, and structural materials originally selected in the
1950’s to develop a nuclear reactor with sufficiently low
mass and high operating temperature to power the jet
engines of a nuclear powered aircraft. The FHR concept
recombines these materials into a new configuration that
uses solid, high temperature fuel and maintains very
high thermal margin to fuel damage while operating at
comparable power density to the original aircraft reactor
design. By producing substantially more power than alternative reactor technologies of similar physical size and
by delivering heat at high average temperature, FHRs have
the potential to have lower capital costs than alternative
reactor technologies.
FHR technology could also support the development of
future reactor technologies that could run more economically on recycled fuel. FHRs may be able to use “deep
burn” fuels already studied for HTGRs, where the FHR
coated particle fuel would fabricated from plutonium and
other actinides. With deep burn fuel FHRs can effectively burn up to 70% of these transuranics. Likewise, the
technology of FHRs closely overlaps that of metal-fueled
fast-spectrum reactors including the IFR, which use fuel
that can be fabricated much more easily from transuranics
than can conventional oxide pellet fuels.
Finally, FHR technology closely overlaps that of the
MSBR. Fluid fuels are easy to produce from recycled spent
fuel, because the production of the fuel involves simple
fluorination and separations processes, and no fabrication
of recycled fuel is required. The Denatured Molten Salt
Reactor [14] provides an example of a technology that
could be used to recycle spent fuel from LWRs and FHRs,
and that could also be used to transition to thorium-based
fuel cycles that would operate sustainably with very low
production of transuranics.
FHRs have the potential to serve as a bridge technology to more advanced reactor technologies. Given the
current and anticipated low production costs for nuclear
energy, today the key development issue for advanced
reactors is to achieve sufficiently low capital cost to compete with LWRs and thus to increase the competitiveness
of fission energy compared to its fossil alternatives.
PET ERSON / F HRs 27
5.0 References
1. Nuclear Energy Institute, “U.S. Electricity Production
Costs in 2010,” http://www.nei.org/filefolder/US_
Electricity_Production_Costs.ppt.
2. J. Uhlir, “Chemistry and technology of Molten Salt
Reactors – history and perspectives,” Journal of
Nuclear Materials, Vol. 360, 6–11(2007).
3. H. G. MacPherson, “The Molten Salt Adventure,”
Nuclear Science and Engineering, Vol. 90, 374-380
(1985).
14. J.R. Engel, et al., “Conceptual Design Characteristics
of a Denatured Molten Salt Reactor With OnceThrough Refueling,” Oak Ridge National Laboratory,
ORNL/TM-7207, July 1980.
15. J.H. Shaffer, “Preparation and Handling of Salt Mixtures for the Molten Salt Reactor Experiment,” Oak
Ridge National Laboratory, ORNL-4616, January
1971.
16.“The Development Status of Molten-Salt Breeder
Reactors,” ORNL-4812, August 1972.
4.http://energyfromthorium.com/msrp/
5. C.W. Forsberg, P. Pickard, and P.F. Peterson, “Molten-Salt-Cooled Advanced High-Temperature Reactor
for Production of Hydrogen and Electricity,” Nuclear
Technology Vol. 144, pp. 289-302, 2003.
6. D.T. Ingersoll et al., “Status of Preconceptual Design
of the Advanced High-Temperature Reactor (AHTR),”
Oak Ridge National Laboratory, ORNL/TM2004/104, May 2004.
7. D. Caron et al., “A Modular Pebble-Bed Advanced
High Temperature Reactor,” NE-170 Senior Design
Project, U.C. Berkeley Thermal Hydraulics Laboratory, Report UCBTH-08-001, May 16, 2008.
8. “NGNP Fuel Qualification White Paper,” Idaho National Laboratory, INL/EXT-10-17686, Rev. 0, July
2010.
9. “NGNP High Temperature Materials White Paper,”
Idaho National Laboratory, INL/EXT-09-17187, June
2010.
10. R. Sims and J. Nestell, Roadmap to Develop ASME
Code Rules for the Construction of High Temperature
Gas-Cooled Reactors (HTGRS), STP-NU-045 Rev. 1,
Feb. 2012.
11. N. Zuber, “An Integrated Structure and Scaling Methodology for Severe Accident Technical Issue Resolution, Appendix D,” U.S. Nuclear Regulatory Commission, NUREG/CR-5809, 1991.
12. P. Bardet and P.F. Peterson, “Options for Scaled Experiments for High Temperature Liquid Salt and
Helium Fluid Mechanics and Convective Heat Transfer,” Nuclear Technology, Vol. 163, pp. 344 – 357,
2008.
13. T. Cisneros, M. Laufer, and R. Scarlat, “Preliminary
Fluoride Salt-Cooled High Temperature Reactor
(FHR) Subsystems Definition, Functional Requirement Definition and Licensing Basis Event (LBE)
Identification White Paper,” Department of Nuclear
Engineering, U.C. Berkeley, Report UCBTH-12-001,
February, 2012.
28
P E T E R S O N / FHRs
Appendix:
Historical Origins of the FHR
The historical origins of the FHR date back to the earliest
phases of nuclear energy development, shortly after World
War II when Ed Bettis and Ray Briant of Oak Ridge National Laboratory (ORNL) were placed in charge of a
major, classified program to design a nuclear-powered
aircraft, which would use a nuclear reactor to heat air hot
enough to drive jet engines. Obviously, for an aircraft to
fly with a nuclear reactor on board, this reactor system
would have to have very low total mass and deliver heat
at a sufficiently high temperature to drive gas turbine jet
engines.
For the new Aircraft Nuclear Propulsion (ANP) program,
Bettis and Briant settled upon an elegant solution to design
an aircraft nuclear reactor. First, they selected a mixture
of fluoride salts to serve as the coolant to transfer heat
from the reactor. The fluoride salts were unique among
the potential coolants due to their chemical inertness and
the very large amount of heat these salts can absorb with
a small change in temperature (greater than water and
over four times larger than for sodium). Likewise, their
very high boiling temperatures, above 1350°C (2400°F),
allowed the reactors to operate at atmospheric pressure.
With this high volumetric heat capacity, these salts could
transfer a large amount of thermal power with a very low
flow rate, allowing the use of small, thin-walled pipes and
compact, light-weight pumps to move the fluid. In their
design the fluoride salt coolant also served as a solvent
for the uranium fuel.
To enable the reactor to achieve criticality and operate
at high power with a very small amount of uranium fuel,
Bettis and Briant selected the ceramic material beryllium
oxide as a “moderator” for the reactor, knowing that the
beryllium atoms are light and can effectively slow down
neutrons during collisions, making it much easier for
subsequent fission reactions to occur with only a small
quantity of uranium fuel.
This combination of fluoride salts, uranium fuel, ceramic core materials, and a high-nickel alloy reactor vessel and primary loop components enabled the design of a
reactor (shown in Fig. 1) that was sufficiently lightweight
that an airplane could lift it, and that could deliver heat
at a sufficiently high temperature that it could power gas
turbine engines needed to maintain flight (the design also
allowed for supplemental jet fuel injection to facilitate
take-off). And such a reactor was built: in 1954 the Aircraft
Reactor Experiment (ARE) operated for 1000 hours delivering 2.5 MWth of power at a temperature of 860°C
(1500°F) [2, 3].
With the launch of Sputnik in 1957 it became clear that
missiles would soon be able to deliver nuclear explosives
to targets over intercontinental distances in under an hour.
So U.S. nuclear deterrence strategy evolved to focus on
missiles rather than bombers, and in 1961 President Kennedy cancelled the ANP program.
In hindsight, the low capital costs of aero-derived gas
turbines and the larger stationary gas turbines that followed them seems to be an obvious result of miniaturization and weight reduction (as with the miniaturization of
electronics and creation of integrated circuits that also
emerged from aerospace applications). Conversely, the
most important military application for nuclear power
ended up being the nuclear powered submarine, where
mass did not matter to the crash reactor-development
program launched in 1948 with the creation of the Navy’s
Nuclear Power Branch. Without the stringent weight
constraints and high temperature requirements that governed the subsequent ANP program, water-cooled reactors,
operating with high pressure in thick-walled reactor vessels and with low thermal efficiency, emerged as the
preferred technology for naval propulsion.
the U.S. in all of its earliest reactors to slow down (or
“thermalize”) the high-energy neutrons emitted from fission reactions. The compatibility of fluoride salts with
graphite was a major factor in its selection as a structural and moderator material in the new MSBR project at
ORNL, because graphite is a robust high-temperature,
low-density ceramic material, capable of sustaining temperatures up to 2000°C without losing mechanical integrity. Used as a structural material in a high-power-density reactor core where local temperatures might become
very high (and in the 1950’s and 1960’s the ability to
predict these temperatures was quite rudimentary), the
MSBR designers could have confidence that the reactor
core would not melt even in the hottest spots and would
not lose its structural integrity during high power operation.
Under the leadership of Alvin Weinberg, the Molten
Salt Reactor Experiment (MSRE) was approved and design
started in the summer of 1960 at ORNL (Fig. 4). The MSRE
core, shown in Fig. 5, was a cylindrical assembly of graphite fuel channels measuring 1.37-m-diameter and 1.62-mhigh in order to minimize neutron leakage. The core had
FIGURE 4
Technicians working on the graphite core of the
Molten Salt Reactor Experiment. The core stood 1.62 m in
height and 1.37 m in diameter [image from Oak Ridge
National Laboratory].
Effort to develop a molten salt fueled breeder reactor
for commercial applications continued at ORNL, along
with larger parallel efforts to develop liquid sodium cooled
fast-spectrum breeder reactors at Argonne National
Laboratory, continued after the cancelation of the ANP
program. Because large-scale industrial experience existed with the use of molten fluoride salts in the electrochemical production of aluminum used in the war, it was
also well known that these fluids could be compatible with
a key high-temperature ceramic structural materials,
graphite, used as the electrode material in the production
of aluminum), as well as with high nickel structural alloys
that had been developed by the ANP project for reactor
vessels and heat exchangers.
Graphite was already very well known in nuclear reactor applications, because it was also the material used by
PET ERSON / F HRs 29
FIGURE 5
On the left, Alvin Weinberg at the control panel of the Molton Salt Reactor Experiment in October 1967 after it
had operated 6000 full power hours at full power. On the right, then AEC Chairman Glenn Seaborg operates the controls
of the Molten Salt Reactor on October 8, 1968 [both images from Oak Ridge National Laboratory].
a power of 8 MWth, which allowed it to still be classified
as an experimental reactor.
Ultimately, an 8-MWth MSRE was built for just over
$8 Million (1961 $) [3] and took approximately 3 years to
construct. The initial fuel for the MSRE was 7LiF-BeF2ZrF4-UF4 [4] while the intermediate coolant was clean7LiFBeF2. In 1968, the original fuel was replaced with 233U
4
making it the first reactor to run on with this fissile fuel .
The moderator used was graphite where the metallic
structural materials selected was primarily Alloy N. The
MSRE ran from 1965-1969 at a typical operating temperature of 600˚C [15]. The intermediate loop of MSRE
used a clean salt without nuclear fuel, as would be used
in FHRs, and experienced no detectable corrosion after
over 26,000 hours of operation [16]5.
Work on the MSBR was terminated in 1972 following
a national decision to focus resources into development
of the LMFBR. Subsequent U.S. and international efforts
to develop high temperature reactor technologies (particularly in Germany) focused on high-pressure helium
cooled reactors using ceramic fuels. LWR technology
derived from the successful application in naval propulsion
emerged as the dominant technology for commercial reactors, and remains so today. But with nuclear power expansion now challenged due to high construction costs and
long construction schedules, a natural question emerges:
might an aero-derived nuclear power plant might be able
to achieve significantly lower construction costs?
30
P E T E R S O N / FHRs
Footnotes
1 There are currently four AP-1000 units under construction at the Sanmen and Haiyang sites in China. Domestically, the NRC recently issue a
combined construction and operating license (COL) to the Southern
Nuclear Operating Company for two 1100-MW AP-1000 model reactors
to be built at the Vogtle site. The NRC is currently reviewing the ESBWR
and final design certification is set for the spring of 2012.
2 Reactors are designed to respond safely to Anticipated Operational
Occurrences with frequencies greater than once every 100 years, Design Basis Events with frequencies greater than once every 10,000
years, and Beyond Design Basis Events (BDBEs) frequencies greater
than once every 5 million years. Because the damage caused by
BDBEs may be difficult to predict, modern designs provide for flexible
and diverse methods to mitigate damage, including the ability to hook
up portable coolant injection and power generation equipment.
3 There are multiple alternative methods that can be used to enrich lithium. Substantial incentives exist to develop domestic U.S. capability
because Li-7 is also needed for chemistry control in pressurized water
reactors. The 69 U.S. pressurized water reactors require a supply of 400
kg per year of Li-7 to remain operational, which is currently purchased
from China where it is produced using the environmentally hazardous
mercury-based processes.
4 Thorium is of interest as potential reactor fuel due to its abundance (4
times greater than uranium) and the very low production of transuranic
elements (plutonium and heavier elements) in the thorium fuel cycle.
The primary fuel for the thorium cycle is U-233, produced by the capture of neutrons in Th-232. FHRs can be designed to produce approximately 30% of their power from thorium, but with fluid fuel, reactor
designs that use thorium exclusively are possible.
5 During operation the concentrations of CrF2 in the fuel salt were observed to rise by a level indicating an average corrosion rate of 4 mills
per year, and after shutdown it was found that fission products had
caused intergranular attack. Because development was stopped, future
deployment of fluid fueled MSRs would require further work to qualify
corrosion resistant materials for use with fuel salts.
Authors’ Biographies
THEODORE U. MARSTON, PhD.
Principal, Marston Consulting
www.marstonconsulting.net
Upon retirement from the Electric Power Research Institute (EPRI) as its Chief Technology Officer, Dr. Marston
established Marston Consulting in June of 2006. This firm is dedicated to the innovation, development, demonstration and deployment of new technology to address two key issues facing developed and developing
countries in the 21st century: energy independence and management of global climate change. His clients
include venture capitalists, commercial and energy companies, R&D organizations, and U.S. and international national laboratories.
Previously, as CTO of EPRI, he directed multi-hundred million dollar, international science and technology
programs to improve the generation, transmission, distribution and utilization of electricity and reduce the
associated environmental risks. Earlier, he led a large international program to develop utility requirements for
advanced nuclear reactors, design certification for advanced light water reactors, first-of-a-kind engineering and siting of nuclear reactors.
In addition to his nuclear experience, he developed international, independent, fossil-fueled power generation projects. Ted has over 30 years
of global experience in the assessment and management of risk in a broad range of industrial facilities, including nuclear and conventional
power plants, refineries, chemical plants, railroads, and defense facilities.
He received his PhD in Mechanical Engineering from the University of Michigan in 1972 and is a Fellow of the American Society of
Mechanical Engineers.
DR. ANDREW C. KADAK
Research Scientist and Former Professor of the Practice,
Massachusetts Institute of Technology, Cambridge, MA
Dr. Kadak is a former Professor of the Practice in the Nuclear Science and Engineering Department at the
Massachusetts Institute of Technology (MIT). His research interests include the development of advanced
reactors, in particular the high-temperature pebble-bed gas reactor, space nuclear power systems, improved
technology-neutral licensing standards for advanced reactors, and operation and management issues of existing nuclear power plants. Before joining the faculty of MIT, Dr. Kadak was principal physicist for pressurizedwater reactor physics at Combustion Engineering Corporation from 1972 to 1975, and held various positions
at Yankee Atomic Electric Company from 1979 to 1987, including president and chief executive officer. He has
served as a board and executive committee member of the Nuclear Energy Institute and the industry’s Advisory Committee on High-Level Waste. He also has served as a member of the National Association of Regulatory Utility Commissioners special panel on high-level nuclear waste and the Aspen Institute’s Dialogue on Nuclear Waste Disposal. In 1995,
Dr. Kadak was a member of the Advisory Committee on External Regulation of DOE Nuclear Safety for the U.S. Department of Energy. He
was president of the American Nuclear Society from 1999 to 2000. Dr. Kadak also serves as a Presidential Appointee on the US Nuclear Waste
Technology Review Board overseeing the Department of Energy’s nuclear waste disposal program.
Dr. Kadak earned a bachelor’s degree in mechanical engineering from Union College in 1967, a master’s degree in nuclear engineering
from the Massachusetts Institute of Technology in 1970, a Ph.D. in nuclear engineering from MIT in 1972, and an MBA from Northeastern
University in 1983.
DR. PER PETERSON
Professor and Chair, Department of Nuclear Engineering, UC Berkeley
Per F. Peterson is Professor and Chair of the Department of Nuclear Engineering at the University of California,
Berkeley. After joining UCB in 1990 as an Assistant Professor he received a NSF Presidential Young Investigator award, and most recently received the Fusion Power Associates Excellence in Fusion Engineering Award
and the American Nuclear Society Thermal Hydraulics Division Technical Achievement Award. His research
interests focus on topics in heat and mass transfer, fluid dynamics, and phase change. He has worked on
problems in energy and environmental systems, including advanced reactors, inertial fusion energy, high level
nuclear waste processing, and nuclear materials management, with over 100 publications on these topics. His
research group currently focuses on high-temperature heat transport for nuclear hydrogen and gas-cycle
electricity production with high temperature reactors, and on the applications of liquid salt coolants in advanced
nuclear energy systems. His broader interests in energy and environment focus on nuclear technologies and
nuclear materials management, and methods to enhance long-term proliferation resistance and physical security in the nuclear fuel cycle.
Dr. Peterson holds a B.S. in Mechanical Engineering, from the University of Nevada, Reno (1982), an M.S. in Mechanical Engineering from
the University of California, Berkeley (1986), and a Ph.D. in Mechanical Engineering from the University of California, Berkeley (1988).
AU T H ORS’ BIOGRAPH IES 31
Notes
This page intentionally left blank.